Appendix 2.7

Engine instrumentation

Layout

The design of engine instrumentation on multi-engined aircraft is inevitably a matter of compromise. The conventional and ergonomically accepted layout is for all instruments associated with a particular engine to be organised in a column, and for all instruments of the same type to be organised in a row. It is, moreover, clearly preferable for each column of instruments to be associated spatially with the throttle of the appropriate engine. This is the basic layout illustrated in Figure 1 and the desirability of using such a layout for the primary engine instruments is clear. Secondary engine information is not required on the front panel of the flight deck in those aircraft with three man crews, and the ideal layout of front panel engine instrumentation described above may thus he adopted.

FIGURE 1

If the aircraft is provided with only two crew members, however, then the secondary engine instruments must be accommodated on the front panel as well. They cannot be accommodated by extending the height of the columns since panel height precludes such an option if the instruments are to be large enough to remain legible.

If the instruments are all to be located on the front panel, two possibilities are apparent. The first is to mount the secondary instruments to one side of the primary instruments as in Figure 2.

Primary	Secondary
Instruments	Instruments
Nol No2	Nol No2

FIGURE 2

The second is to split the secondary instruments and mount them outboard of their respective primary instruments, as in Figure 3.

No 1 engine

No 2 engine

FIGURE 3

The advantage of the layout in Figure 3 is that the instruments for a given engine are all mounted together and are, if not spatially, at least cognitively, aligned with their associated power levers. This is achieved at the price of splitting the secondary instruments apart, with the associated possibility of disparate secondary readings going undetected.

Figure 2 achieves the goal of keeping the instruments paired together, and thus maximises the chances of disparate readings being detected, but does so at the price of splitting up the instruments associated with a given engine, and of losing the advantage of having all instruments cognitively aligned with their corresponding throttle levers.

Thus, Figure 3 could fairly be judged to maximise the probability that a given failure will be correctly identified by the crew as belonging to a given engine, at the possible cost of less efficient error detection on the secondary instruments, whereas Figure 2 may be judged as maximising the probability that disparate readings will be detected at the cost of degrading the probability that this detected failure will be associated by the crew with the correct engine.

The design of the EIS

The layout of the EIS in the Boeing 737 Series 300/400 conforms to Figure 2, which has been widely used without apparent difficulty in many two-engined, two-pilot aircraft. The illumination of the display, however, might aggravate the problem of perceived misalignment of the instruments with their respective throttles. On the hybrid instruments (LED counters with electromechanical pointers) fitted to other aircraft of this type, the faces of the instruments needed to be lit from in front to show the pointers, dials and scale marks. Such lighting does not, of course, illuminate only the legends and pointers on the instruments but also the general structure and limits of the display, so that the instruments could be argued to be viewed within a structured visual frame. In the EIS display, all symbology is edge-lit and set against a heavily constrasting background which, in an aircraft at night will be, to all intents and purposes, black. This may have the effect of enhancing the extent to which the instruments are seen as a single display rather than as two separate displays, and may degrade the extent to which deviant readings in, say columns 1 and 3 of the matrix could readily he associated with the No 1 engine.

The next most obvious and important change made between the hybrid system and the EIS is that the full-radius mechanical pointers have been changed to short LED pointers moving round the outsides of their scales. The mechanical pointers were relatively large, white and clearly linear devices, and their orientation on the display was immediately apparent. Not only was the absolute orientation of each pointer apparent but (and perhaps more importantly) it was readily apparent whether the pointers of each pair of instruments were parallel with one another. The pointers on the LED display are much shorter than the mechanical pointers, they are the same colour as the LED counters and they move in steps. They are much less conspicuous than the mechanical pointers, acting more as scale markers, and providing less immediate directional information. They are thus less well able to give the comparative information provided by the strong cue of parallelism of the mechanical pointers. This comparative information can be obtained with certainty only by interrogating each instrument to see if the LED pointers of each pair are at the same points on the scale or by comparing the readings of the pairs of counters.

Evaluatton and testing

The entire function of any display on a flight deck is to transfer information from the aircraft to the pilot, and to do so in the way that will cause the pilot least workload and will be least likely to be interpreted wrongly. Although some principles, such as those discussed above, guide the design of displays, the only way of evaluating the adequacy of a display is by experiment and trial. It is therefore important that before any display is put into service, it is subjected not just to some form of acceptability judgement by company pilots, but to a structured assessment using average line pilots. Indeed, it could he argued that such assessments should he conducted using the least able pilots who are ever likely to use the display.

A display similar to the EIS was developed by Smith's Industries for use on the McDonnell Douglas MD88. It was held to differ from an earlier display which employed mechanical pointers, in that the colour coding of some dials was changed. The new display was evaluated by pilots employed by McDonnell Douglas and the Federal Aviation Administration (FAA). The evaluation was held to show that the new display provided clearly readable and interpretable information to the flight crew, showed whether the current state of power plant operation was normal or abnormal, indicated the engine maximum/minimum safe operating range and showed whether the system(s) operation was being accomplished in a safe manner. These results were used by McDonnell Douglas to demonstrate to the FAA the acceptability of the new display as an equivalent means of compliance with current airworthiness regulations.

The EIS for the Boeing 737 was designed to represent a minimum change from the previous hybrid display and, accordingly, it was type certified by both the FAA and the CAA as fit for its purpose. The counters remained identical in size and colour but the dials of all instruments were reduced in size. The pointers were reduced in length by approximately two-thirds and placed on the outsides of the dials but the circumference swept by the needle tips (ie the instrument 'size') remained the same. The EIS display was deemed to have sufficient communality with the hybrid display to circumvent the need for pilots to be separately rated for EIS-equipped models. It was tested for proper operation, compatibility and freedom from electrical interference but it was not evaluated for its efficiency in imparting information to pilots.

Although the desire for communality is understandable, because a number of other factors were changed between the hybrid and the EIS displays, the apparent benefit of keeping size constant may have been offset or even negated by varying others such as illumination, contrast and pointer size. The desire to maintain consistency of di splay format while introducing new technology was responsible for the reduction in pointer size and con spicuity, and exemplifies a general problem. LED and CRT displays possess potential advantages over old technology instrumentation that may be exploited only if the display is designed afresh to exploit them. If a new technology display is designed simply to mimic the appearance of its precursors it may well fall into what is sometimes referred to as the 'electric horse' trap; the strengths of the old system are discarded because they cannot be duplicated, and the potential strengths of the new system are not exploited. Full length pointers cannot he represented on the LED system because the packing density of central LEDs cannot be achieved, and because symbology cannot be overlaid, and a potentially less satisfactory pointer is substituted.

It is reiterated that the general effectiveness of any new display may be judged only by trial and experiment, but even then some criterion of acceptability must be adopted. An obvious criterion in the case of engine instrumentation is that the new display should not prove less satisfactory to those pilots who use it than the display it replaces. When the EIS was introduced for use on the Boeing 737 no such tests were carried out.

Conclusions

Although there seems to be no question that the EIS display on the Boeing 737 provides accurate and reliable information to the crew, the overall layout of the displays, and the detailed implications of small LED pointers rather than the larger mechanical ones, and of edge-lit rather than reflective symbology do appear to require further consideration. These factors should not be ignored and the suitability of such new displays for use by airline pilots should be evaluated before they are brought into use.

This page was copied from:	http://www.open.gov.uk/aaib/gobme/a27.htm
COPY! Last modification on 1999-06-15 by Michael Blume